Retina Segmentation to Obtain the Optic Nerve Area and Blood Vessels


Abstract in English

This paper proposes a new approach for the segmentation of the retina images to obtain the optic nerve and blood vessels regions. We used retinal images from DRIVE and STARE databases which include different situations like illumination variations, different optic nerve positions (left, right and center). Illumination problem has been solved by preprocessing stage including image histogram-based illumination correction. Next, some morphological operations were used to filter the preprocessed image to obtain the ROI region, then, the center and radius of optic nerve were determined, and the optic nerve region was extracted from the original image. In blood vessels segmentation, we applied the illumination correction and median filtering.Then the closing, subtraction and morphological operations were done to get the blood vessels image which was thresholded and thinned to get the final blood vessels image.

References used

LALONDE, M, “Fast and robust optic disc detection using pyramidal decomposition and Hausdorff-based template matching”, IEEE Trans. Med. Imaging, 2001
HAAR, T. F., “Automatic localization of the optic disc in digital color images of human retinal”, Utrecht University, 2005
R.ANAND, G.ARAVINTH BABU, K.DEEPA and V.LEKHA, “SEGMENTATION OF OPTIC DISC INFUNDUS IMAGES”, Indian Journal of Computer Science and Engineering, Vol. 3, 2012, pp.230-234

Download